Joseph C. Corbo - US grants

DESCRIPTION (provided by applicant): Understanding the mechanistic basis of retinal disease will require elucidation of the transcriptional regulatory networks (TRNs) which control gene expression during retinal development. Recent studies have shown that mutations in several transcription factors (TFs) controlling photoreceptor cell fate and differentiation underlie retinal disease in humans and mice. One such TF, Crx, not only causes cone-rod dystrophy when mutated, but a number of its downstream target genes have been implicated in retinal disease. In addition, a second TF required for normal rod differentiation, NRL, has been implicated in retinitis pigmentosa. The target genes of this latter TF largely remain to be identified. It is hypothesized that Crx and NRL lie near the top of a hierarchical TRN controlling rod photoreceptor differentiation. The overall goal of this proposal is to analyze the gene network controlled by the TFs, Crx and NRL, during mammalian rod photoreceptor development. These studies will serve to define the TRN topology of the rod cell, thereby laying the groundwork for future systems-level analyses of gene function in photoreceptor development and disease. Specific aim 1 will identify the target genes controlled by NRL as well as those coordinately regulated by Crx and NRL using microarray technology. Specific aim 2 will locate the putative cis-regulatory elements of these target genes employing bioinformatic approaches and test their function in vivo via electroporation of enhancer/reporter fusion constructs. Specific aim 3 will determine which of the putative target genes are direct targets of Crx and NRL by microarray analysis of retinae carrying transgenes overexpressing Crx and NRL in the presence or absence of protein-synthesis inhibitors. Specific aim 4 will determine the subsets of target genes controlled by TFs downstream of Crx and NRL using microarray analysis. Dr. Joseph Corbo, the Principal Investigator, is an M.D., Ph.D. with a doctoral degree in Biology and residency training in Anatomic Pathology and Neuropathology. He now seeks further training in genomics and cell biology under the mentorship of Dr. Constance L. Cepko, whose lab studies cell fate determination and differentiation in the retina. The candidate will combine these newly acquired skills with his clinical training in Neuropathology to pursue a career as a clinician-scientist applying cutting edge genomic technology to systems-level analyses of development and disease in the mammalian eye and brain.

[unreadable] DESCRIPTION (provided by applicant): Photoreceptors are subject to a greater number of Mendelian diseases than any other cell type in the human body. Unfortunately, the repertoire of cis-regulatory elements (CREs; i.e., promoters/enhancers) available for targeting gene therapy to photoreceptors is extremely limited. It is our aim to develop a quantitative understanding of photoreceptor CRE function that will facilitate the identification of novel photoreceptor-specific CREs for gene therapy and will inform the rational engineering of artificial gene circuits for therapeutic purposes in photoreceptors. Achieving these goals will require a detailed understanding of the cis-regulatory networks that control gene expression in photoreceptors. Accordingly, we have produced a comprehensive model of the photoreceptor transcriptional network controlled by the transcription factors (TFs), Crx and Nrl. In the course of this work, we developed a computational algorithm, Phastfind, to predict CREs around hundreds of genes in this network. We then created a high throughput validation pipeline to assay CRE activity in living retinas. This assay has so far led to the identification of 19 novel CREs around retinal disease gene loci, thus doubling the number currently available for gene therapy. The present proposal aims to extend this newly gained knowledge of photoreceptor cis-regulation by further elucidating the role of Crx and Nrl in controlling photoreceptor CRE activity and by exploiting these two key transcriptional regulators for therapeutic purposes. We hypothesize that the affinity, spacing and orientation of Crx and Nrl sites within a photoreceptor CRE quantitatively control its transcriptional activity in a predictable fashion. We will test this hypothesis in Specific Aim #1 by systematically elucidating the quantitative contributions of Crx and Nrl binding sites to transcriptional activity in both natural and synthetic CREs. Next, we will apply our knowledge of Crx and Nrl for therapeutic purposes in the retina. In Specific Aim #2 we will use the results of the Phastfind algorithm to create a `minimalized' gene-specific CRE for Crx and use it in a gene therapy vector to treat a mouse model of congenital blindness. This Aim will serve as test case for a general approach to CRE design that combines computational prediction with rapid in vivo validation. If this approach is successful, we believe it can be used to engineer compact, vector-ready gene-specific CREs for a wide range of human retinal disease genes. In Specific Aim #3 we will exploit Nrl's role as a determinant of rod cell fate to engineer a synthetic drug-inducible cell fate switch which can be used to alter the fate of developing photoreceptors for therapeutic purposes. We hypothesize that this switch may permit treatment of a wide range of diseases caused by mutations in rod-specific genes by driving the transdifferentiation of diseased rods into cones. In addition, this switch could someday be used to regulate the differentiation of embryonic stem cells into photoreceptors for replacement therapy. Overall, the proposed studies promise to deliver tools that can be directly translated into clinical therapies for patients with blindness. PUBLIC HEALTH RELEVANCE: Photoreceptors in the retina are the main cell type affected in patients with blindness. Unfortunately, the repertoire of photoreceptor-specific promoters used to make gene therapy vectors to treat these patients is very limited. It is the aim of our research to significantly expand the repertoire of both natural promoters and synthetic gene circuits available for treating patients with blindness. [unreadable] [unreadable] [unreadable]

DESCRIPTION (provided by applicant): 98% of the mammalian genome is non-coding, a significant fraction of which has cis- regulatory function. If we are ever to interpret the effects of the vast number of non-coding polymorphisms in the human population, we must gain a more complete understanding of the cis-regulatory grammar of gene expression. Fueled by genomic-scale technologies, the identification of potential cis-regulatory elements (CREs) is proceeding rapidly. Unfortunately, the rate of discovery of putative CREs far outpaces our capacity to functionally analyze them in vivo. Clearly, there is an urgent need for new high-throughput technologies to elucidate the functional architecture of CREs in mammalian cells. Reporter gene analysis is the standard assay for dissecting the function of mammalian CREs, but the time- and labor-intensive nature of the assay makes it impractical for testing large numbers of CREs. We propose to harness the power of Next Generation Sequencing (NGS) to create a system for the multiplexed analysis of reporter genes in mammalian cells. Our strategy is to fuse libraries of CREs to barcoded reporter genes and quantify their output by NGS. We will demonstrate the utility of this assay by studying cis-regulation in a mammalian neuronal cell type, the retinal photoreceptor cell. In Aim 1 we will dissect, at single nucleotide resolution, the photoreceptor- specific Rhodopsin (Rho) promoter. More than 15,000 mutant promoters will be assayed in a single, multiplexed assay in living retinas, an experiment that is not possible with any existing methodologies. The data from this experiment will allow us, for the first time, to quantify the relationship between evolutionary sequence conservation and cis-regulatory activity at nucleotide resolution. In Aim 2 we will deploy the assay to test 1,000 Chip-seq "peaks" that are enriched for binding to the key photoreceptor transcription factor, Crx. In a single, multiplexed experiment we will test all 1,000 genomic sequences for their ability to drive photoreceptor- specific expression in vivo. In addition, in Aim 2 we will assay combinatorial libraries of synthetic promoters composed of binding sites for Crx and two other transcription factors, Nrl and Nr2e3, known to play a key role in controlling photoreceptor-specific gene expression. The output of all these experiments will be analyzed using a formal thermodynamic model of combinatorial cis-regulation. Our intention is to demonstrate, in a relatively short time frame, that it is possible to unravel the cis-regulatory grammar of an important mammalian cell type in vivo. This study will serve as a proof-of-concept that CRE analysis can be made as high- throughput as CRE discovery. PUBLIC HEALTH RELEVANCE: Differences in the DNA sequence of the human genome account for variation between individuals in their susceptibilities to many diseases. Many of these genetic differences occur in regions of the genome containing stretches of DNA that control which genes are turned on and off. The goal of the proposed research is to develop a new methodology for determining the function of these control regions so that we can better predict the effects of human genetic variation on disease susceptibility.

Next-generation sequencing (NGS) technologies such as ChIP-seq are driving the discovery of putative cis-regulatory elements (CREs; i.e., enhancers/promoters) at an astonishing pace. Testing the regulatory potential of these predictions is a serious bottleneck in our efforts to understand the logic of gene regulation in the human genome. Currently there are no viable technologies for functionally testing the thousands of predictions that can be generated from even a single ChIP-seq experiment, let alone the millions of predictions being generated by ENCODE and other consortium-based efforts. Breaking the logjam of CRE predictions requires a new technology that enables massively parallel cis-regulatory analysis in mammalian cells. In this proposal, we introduce CRE-seq (Cis-Regulatory Element analysis by sequencing), a novel technique for assaying thousands of CREs in a single experiment in mammalian cells. Our strategy is to fuse libraries of CREs to barcoded reporter genes, transfect these libraries into cells and quantify their output by NGS. We will develop CRE-seq as an efficient, robust and highly parallel technology for assessing the cis-regulatory activity of thousands of ChIP-seq peaks in a single experiment. If successful, this strategy should make it possible to quantify the promoter activity of entire 'cis-regulomes' (i.e., the entire complement of cis-regulatory regions controlling gene expression in a given cell type). We will demonstrate the utility of this assay by studying cis-regulation in two different mammalian cell types, retinal photoreceptors (a differentiated neuronal cell type) and undifferentiated embryonic stem cells. CRE-seq promises to advance our understanding of human gene regulation and will serve as a novel source of personalized genomic information available for diagnosis and treatment of disease.

DESCRIPTION (provided by applicant): Optogenetics holds tremendous potential for restoring vision to individuals with late-stage retinal degeneration, particularly those patients who have lost most of their photoreceptors. One promising therapeutic strategy is to express a light-sensitive protein in non-photosensitive bipolar cells by gene therapy. Current approaches are limited by inefficient bipolar targeting and expression, and require application of potentially phototoxic levels of blue-green light to stimulate the optogenetic actuator. The objective of the present proposal is to overcome these challenges by utilizing directed evolution and synthetic biology to engineer an AAV-based delivery system to target red-shifted optogenetic devices to both ON and OFF bipolar cells, and to employ this system to treat blindness in mice. In Aim 1, we will use directed evolution to engineer new AAV serotypes capable of highly efficient bipolar AAV infection after injection into the vitreous humor. In Aim 2, we will utilize a novel technolog called CRE-seq to engineer thousands of compact, ON bipolar-specific promoters that exhibit excellent specificity and a wide range of expression strengths. In addition, we will engineer an AAV-deliverable synthetic gene circuit to target an optogenetic inhibitor specifically to OFF bipolar cells. In Aim 3, we will combine the tools developed in Aims 1 and 2 with the use of red-shifted optogenetic devices to restore functional vision to rd1 mutant mice. We recently discovered the enzyme responsible for the 'rhodopsin- porphyropsin' switch in vertebrates, and we will use this enzyme to red-shift optogenetic devices, making them sensitive to far red light (> 650 nm). This therapeutic approach has the potential to dramatically improve light- sensitivity in the rescued mice and will avoid the retinal damage associated with high-intensity blue light exposure, thereby permitting unprecedented levels of functional restoration and setting the stage for future trials in human patients.

? DESCRIPTION (provided by applicant): Cis-regulatory elements (CREs) play a critical role in the regulation of gene expression by mediating the interaction between transcription factors (TFs) and their target genes. Mutations within CREs can disrupt key TF binding sites, thereby altering gene expression, and thus contributing to disease. The vast majority of SNPs identified in genome-wide association studies (GWAS) of complex disease fall within non-coding regions, and is thought to affect the activity of CREs. Thus, a better understanding of the location and function of CREs is essential for interpreting the functional significance of genetic variation within non-coding regions. ENCODE and other consortium-based projects have begun to map the location of CREs in multiple cell lines and whole organs. However, our understanding of cis-regulatory architecture at the level of individual primary cell types is limited. In this proposal,we will utilize a newly developed epigenomic mapping technology, ATAC-seq, to comprehensively identify the CREs of all seven major cell classes in the mouse retina: rods, cones, horizontal cells, bipolar cells, Müller glia, amacrine cells, and ganglion cells. ATAC-seq utilizes a transposase-based approach to tag regions of open chromatin and can be applied to very small numbers of cells purified by fluorescence-activated cell sorting (FACS). This transformative approach will permit us to probe the complex interrelations between chromatin accessibility, nucleosome positioning, and TF binding on a genome-wide scale in all mouse retinal cell classes. Furthermore, we will leverage these data to probe the function of thousands of CREs within specific retinal cell types, using CRE-seq, a novel technique for high-throughput cis-regulatory analysis. Taken together, these studies will generate a comprehensive, functional map of retinal CREs that will serve as a blueprint for understanding the transcriptional networks of the retina and the effects of non-coding variants on disease. In addition, these studies will se the stage for future work that will utilize epigenomic profiling to characterize changes that occur during the process of retinal development and in the course of degeneration.

? DESCRIPTION (provided by applicant): Optogenetic actuators are ion channels or pumps that can be regulated by light, thus permitting neuronal activity to be turned on and off with high spatial and temporal precision. Optogenetics holds significant promise for restoring vision to blind patients, but current treatment strategies require the application of high-intensity blue-green light, which poses a significant risk of retinal photodamage. Dependence on the use of short-wavelength light therefore represents a major barrier to safe and effective implementation of optogenetic therapy for retinal disease. This barrier can be surmounted by the use of optogenetic actuators with red-shifted excitation spectra. Red light is less energetic and therefore less damaging to the retina. Accordingly, researchers have sought to develop red-shifted optogenetic actuators, and considerable progress has been made in red-shifting actuators via opsin engineering. However, in order to extend the operational range of optogenetic actuators into the near-infrared (>700 nm), new orthogonal approaches are needed. The goal of the present proposal is to introduce a novel biomimetic strategy for red-shifting optogenetic actuators: red- shifted chromophore substitution. This approach is complementary to opsin engineering and is based on a strategy used by migrating fish to enable better vision in turbid water. When salmon migrate from the open ocean into inland streams (where incident light is significantly red-shifted), they switch from using retinal as their visual chromophore to , 4-didehydroretinal which has red-shifted spectral properties. This chromophore switch causes a dramatic red-shift of the fish's opsin spectral sensitivity, thereby enhancing the animal's abilityto see long-wavelength light and thus permitting the animal to peer more deeply into turbid streams. Our goal is to identify the enzyme mediating the conversion of retinal into 3, 4-dodehydroretinal, and to co-express it with optogenetic actuators in mammalian neurons in vivo, thereby red-shifting their action spectra. In Specific Aim 1, we use transcriptome profiling in zebrafish and bullfrog to identify the enzyme mediating this conversion, and then characterize its function in vivo. In Aim 2, we will co-express this enzyme with red-shifted optogenetic actuators in vivo to endow non-functioning mouse photoreceptors with sensitivity to near-infrared light (>700 nm). A key feature of this approach is that chromophore substitution can be coupled to the use of any existing actuator in any part of the mammalian nervous system. Thus, this proposal promises to have a widespread impact on the field of optogenetics.

Project Summary One of the major unsolved challenges of the genomic era is to understand how individual sequence variation contributes to disease. Coding variation is increasingly well understood, but our knowledge of the effects of non-coding variation remains rudimentary. The goal of this proposal is to develop a platform for the analysis of non-coding cis-regulatory variation in human retinal disease. We hypothesize that sequence variants in photoreceptor-specific cis-regulatory elements (CREs; e.g., enhancer/promoters) play an important role in retinal disease by altering the expression levels of disease-related genes. Currently, assessing the effects of variants that fall within non-coding DNA represents a challenging problem, in part, because our ability to assay CREs in a high-throughput fashion is limited. To address this challenge, we have developed a technique called CRE-seq (Cis-Regulatory Element analysis by sequencing). In CRE-seq, individual CREs are fused to reporter genes, each containing a unique DNA barcode. The resultant CRE-reporter library, consisting of thousands of constructs, is introduced into living retina, and reporter gene expression is quantified by counting barcoded transcripts with RNA-seq. CRE-seq promises to revolutionize our ability to measure the effects of human cis-regulatory variants. To achieve this goal, we propose three Specific Aims. In Aim 1, we will use ATAC-seq to identify candidate CREs in both developing and mature human photoreceptors. In Aim 2, we will utilize a combination of computational and experimental approaches (including CRE-seq analysis) to analyze the CREs identified in Aim 1 and thereby elucidate the cis-regulatory grammar of human photoreceptors. CRE-seq will be performed in both mouse retinas as well as ES cell-derived human retinal organoids. These studies will provide the first comprehensive view of the human photoreceptor 'cis-regulome' and will begin to decipher the cis-regulatory code of human photoreceptors. In Aim 3, we will use a ?mutagenesis and screening? approach to engineer a library of compact (150 bp), highly active enhancers for targeting human photoreceptors in gene therapy applications. If successful, these studies will establish a quantitative platform for the analysis of non-coding cis-regulatory variation, thereby enabling comprehensive interpretation of whole-genome sequence data in the context of retinal disease. In addition, they will engineer a suite of new enhancers for human retinal gene therapy.

Project Summary Cardiac arrhythmias are a major clinical problem and can predispose to sudden cardiac death. Genome- wide association studies (GWAS) have identified a growing number of sequence variants associated with cardiac arrhythmias and related electrocardiogram (ECG) traits, but the majority of these GWAS hits fall within non- coding regions and their functional effects are difficult to decipher. We hypothesize that the majority of functional non-coding variants related to cardiac arrhythmias fall within cardiac cis-regulatory elements (CREs; i.e., enhancers/promoters), and exert their effects by disrupting transcription factor (TF) binding sites and thereby altering the expression level of genes encoding cardiac proteins, especially ion channels and their regulators. To identify causal variants underlying cardiac arrhythmia-related GWAS hits and to map arrhythmia-related CREs, we propose to implement a technique called CRE-seq (Cis-Regulatory Element analysis by sequencing). In CRE-seq, individual CREs are fused to reporter genes, each containing a unique DNA barcode. The resultant CRE-reporter library, consisting of thousands of constructs, is introduced into living tissue, and reporter gene expression is quantified by counting barcoded transcripts with RNA-seq. CRE-seq promises to greatly accelerate our ability to measure the effects of cis-regulatory variants in cardiac disease. To achieve this goal, we propose two Specific Aims. In Aim 1, we will use CRE-seq to identify causal cis-regulatory variants at all known GWAS loci associated with cardiac arrhythmias and related traits. We will measure the cis-regulatory activity of thousands of wild-type and variant CREs in mouse heart in vivo and in human iPSC-derived cardiomyocytes via adeno-associated virus (AAV)-mediated CRE-seq library delivery. We will then evaluate the functional effects of selected variants on TF binding using protein-microarrays containing all known human TFs. Lastly, we will correlate the results of our CRE-seq analyses with cardiac eQTL data. In Aim 2, we will establish a template for interpreting rare arrhythmia-related variants by mapping the location of human cardiac CREs and elucidating their cis-regulatory logic. We will utilize a 'capture and clone' strategy for CRE-seq library construction, which permits analysis of long (i.e., ~500 bp) tiled reporters at each locus. In this way, we will pinpoint essential TF binding sites (TFBSs) which are the likely targets of rare functional variants. Next, we will use CRE-seq to analyze the effects of introducing all possible single-nucleotide substitutions into identified TFBSs. As in Aim 1, we will perform CRE-seq in both mouse heart and human iPSC-derived cardiomyocytes. Taken together, these two Aims will enable functional interpretation of both common and rare variants in individual human genomes and thereby facilitate assessment of cardiac disease risk in patients.

Project Summary Neuropsychiatric diseases affect millions of people world-wide. Genome-wide association studies (GWAS) have identified a growing number of sequence variants associated with neuropsychiatric diseases and related traits, but the majority of these GWAS hits fall within non-coding regions and their functional effects are difficult to decipher. We hypothesize that the majority of functional non-coding variants related to neuropsychiatric disease fall within brain cis-regulatory elements (CREs; i.e., enhancers/promoters), and exert their effects by disrupting transcription factor (TF) binding sites and thereby altering the expression level of genes encoding proteins expressed in the brain, particularly the cerebral cortex. To identify causal variants underlying neuropsychiatric disease-related GWAS hits and to map neuropsychiatric disease-related CREs, we propose to implement a technique called CRE-seq (Cis-Regulatory Element analysis by sequencing). In CRE-seq, individual CREs are fused to reporter genes, each containing a unique DNA barcode. The resultant CRE-reporter library, consisting of thousands of constructs, is introduced into living tissue, and reporter gene expression is quantified by counting barcoded transcripts with RNA-seq. CRE-seq promises to greatly accelerate our ability to measure the effects of cis-regulatory variants in neuropsychiatric disease. To achieve this goal, we propose two Specific Aims. In Aim 1, we will use CRE-seq to identify causal cis-regulatory variants at all known GWAS loci associated with neuropsychiatric diseases and related traits. We will measure the cis-regulatory activity of thousands of wild-type and variant CREs in mouse cerebral cortex in vivo and in human iPSC-derived forebrain organoids via adeno-associated virus (AAV)-mediated CRE-seq library delivery. We will then evaluate the functional effects of selected variants on TF binding using protein-microarrays containing all known human TFs. Lastly, we will correlate the results of our CRE-seq analyses with brain eQTL data. In Aim 2, we will establish a template for interpreting rare neuropsychiatric disease-related variants by systematically mapping the location of human brain CREs. We will utilize a 'capture and clone' strategy for CRE-seq library construction, which permits analysis of long (i.e., ~500 bp) tiled reporters at each locus. In this way, we will pinpoint essential TF binding sites (TFBSs) which are the likely targets of rare functional variants. Next, we will use CRE-seq to analyze the effects of introducing all possible single-nucleotide substitutions into identified TFBSs. As in Aim 1, we will perform CRE- seq in both mouse brain and human iPSC-derived cerebral organoids. Taken together, these two Aims will enable functional interpretation of both common and rare variants in individual human genomes and thereby facilitate assessment of neuropsychiatric disease risk in patients.